Reassessing the Scottish Mesolithic-Neolithic transition: questions of diet and chronology

The purpose of this research was to broaden our knowledge of the dietary changes and timing of the Mesolithic-Neolithic transition in Scotland. Despite the rich number of archaeological sites around mainland Scotland and the Orkney Isles that date to the transition period (c.4000 to 3600 BC), bones of humans and especially of fauna are rarely recovered. This lack of skeletal material necessary to investigate individual human diet has resulted in a gap in our understanding about dietary changes that occurred during the transition from a hunter-gatherer lifestyle to farming. One of the most widely used and reliable scientific tools available to investigate ancient diet is δ13C and δ15N isotope analysis; however, this technique requires samples of human bone and also a representative sample of bones from animals from the same spatial and chronological context as the humans to form an isotopic baseline. Studies of dietary change during the Scottish Mesolithic-Neolithic transition so far have largely relied on poor quality baselines to interpret the diets of the relatively small number of humans recovered from this period. A solution was therefore required in order to ensure that interpretations of dietary change during this time that employed stable isotope analysis of bone collagen were better defined and more secure. In addition to the problems of interpreting changes in the diet, there are questions surrounding the timing of the Scottish Mesolithic-Neolithic transition as well. The dietary changes were characterised by replacing marine protein with terrestrial animal protein, however we are uncertain as to how fast and complete this shift in diet was. There is long-standing debate regarding how important marine protein was in the early Neolithic diet. If marine protein was only a minor component of the diet, this would be undetected by traditional interpretations of stable isotope data. However, the radiocarbon ages of the bone collagen of these humans may be affected by oceanic 14C. Ancient marine derived carbon in human bone collagen is on average c.400 years older than terrestrial carbon, resulting in the need to correct the ages of mixed marine/terrestrial samples for this Marine Reservoir Effect. Undetected marine protein in the Neolithic diet would result in human bones being assigned radiocarbon dates that are older than the true age of the sample. An improved method of interpreting stable isotope data from bone collagen was therefore sought to address these questions of timing in the Scottish Mesolithic-Neolithic transition. δ13C, δ15N and δ34S isotope analysis was employed on human and faunal bones from Mesolithic and Neolithic sites predominantly in Orkney and the west coast of Scotland. Isotope faunal baselines were supplemented, where appropriate, with modern faunal samples. The abundance of marine and terrestrial resources in the human diet was modelled using the Bayesian mixing model, FRUITS (food reconstruction using isotopic transferred signals). Where marine protein was detected in the human diet, the radiocarbon dates of these samples were calibrated using the mixed marine/terrestrial radiocarbon calibration curve and compared against previously calculated dates, where these were available. It was possible to supplement marine faunal isotope baselines with modern marine samples; however, modern terrestrial samples were unsuitable analogues. δ34S was also found to be an unsuitable proxy for diet in this research. Small amounts of dietary marine protein were detected in the majority of Neolithic humans; however, marine consumption did not have a significant effect on the radiocarbon dates of these individuals. The key finding of this research was, therefore, that the transition from the Mesolithic lifestyle to the Neolithic in Scotland was a lengthy and gradual process, contributing to the debate regarding the nature of the transition. The chronology of the transition that has been previously established by radiocarbon dating is secure: while modelling the isotope data in FRUITS resulted, in most cases, in a greater age range in calibrated radiocarbon dates, they were not erroneously old, as initially predicted